Overview
Meridian Tower is a five-participant interactive installation about how collective failure emerges from locally situated decisions. Each participant occupies a different institutional role with partial information and unequal authority. The system registers their decisions in a shared state, allowing one participant’s action to change the conditions another later encounters.
Private headphone audio and spatial sound shape each participant’s access to the unfolding event. Networked interfaces register their responses within the shared system. Projection reveals changes in the collective situation and makes the distribution of responsibility increasingly perceptible.
March 2026 Playtest — Immersive Media Lab, Qualcomm Institute, UC San Diego
Research Questions
A. Multi-Participant Distributed Authorship
How can a multi-participant computational narrative use distributed agency to reflect how responsibility becomes diffused across social systems, producing collective failure?
B. Information Asymmetry
How can game mechanics stage the tension between private channels and a shared public soundscape, turning information asymmetry into a narrative force?
C. Spatial Agency
How can a real-time audiovisual spatial system translate participant agency into changes in the environment that produce divergent narrative outcomes?
D. Human-System Co-Authorship
How can a computational system participate in narrative authorship by interpreting participant actions and transforming them into new narrative conditions?
HOW IT WORKS
Emergent Narrative & Drama Management
Meridian Tower uses a fixed five-act progression as its dramatic frame. Within this structure, participant decisions are registered in a shared state. Each update changes the narrative conditions available to other roles, allowing the causal structure of each run to emerge through interaction.
The system manages dramatic progression through these state changes. Publication order shifts the public account of the crisis. Later decisions alter the available responses and determine how the event develops. Each run therefore produces a distinct causal history inside the same authored dramatic structure.
SOUND AS OPERATIVE MEDIA
1. Private headphone audio functions as input.
It delivers role-specific fragments of evidence, giving each participant access to a different partial reality.
2. Published sound enters the loudspeaker system as participant output.
Participants filter that evidence through their interests and authority before deciding what enters the shared acoustic space. These selective disclosures construct an evolving public soundscape.
3. The public soundscape is highly spatialized and continuously recomposed in real time by the state machine.
Each participant action updates the current mix, reshaping the presence, density, and spatial behavior of its compositional layers.
4. Sound makes accumulated responsibility perceptible.
The evolving soundscape carries traces of earlier decisions, revealing how distributed agency develops into collective consequence over time.
Spatial Environment & Technical Implementation
Four Projection Surfaces
Left, center, right, and floor projections extend the fictional architecture of Meridian Tower into the physical room. Visual content responds to the shared state in real time, shifting focus and altering the spatial conditions around the participants.
Spatial Audio
A 32-channel loudspeaker system surrounds the participants in 360°, with speakers also positioned at floor level. Sound moves around and beneath the audience, allowing changes in the game state to take on a physical position in the room. The spatial mix evolves with each run and becomes an audible record of the system’s changing state.
Interactive System
Max/MSP · Node.js · OSC/UDP · WebSocket · TouchDesigner · HTML interfaces · Five networked iPads
Research & Writing
Sound-Operative Narrative Systems (SONS)
Meridian Tower develops SONS as a framework for multi-participant interactive narrative. Participants act through a shared narrative state, where one participant’s registered action can alter the later conditions encountered by another. Sound becomes operative when it participates in this transfer of consequence.
ICIDS 2026 Late-Breaking Work
Shi, S. J. (2026). Composing Consequence: Sound as an Operative Narrative System in Multi-Participant Interactive Digital Narrative. 19th International Conference on Interactive Digital Storytelling (ICIDS 2026), Late-Breaking Work.
Accepted · Forthcoming in Springer proceedings (pending final publication).
_edited.jpg)
ITERATIONS
December 2025
UC San Diego · Experimental Theater
62-speaker prototype
March 2026
Qualcomm Institute · Immersive Media Lab
32-channel spatial audio iteration
October 2026 — Upcoming
Qualcomm Institute · Calit2 Theater
Selected Project, IDEAS — Initiative for Digital Exploration of Arts and Sciences
Peer-reviewed interdisciplinary artist residency and public presentation
Credits
Concept, Narrative & Interaction Design: Scarlett Shi
Sound Design & Composition: Scarlett Shi
Visual Design & Projection: Lindsy Gu
Interaction Programming, Shared-State System & Networking: Ruiqi Chu, Scarlett Shi
Max/MSP Programming & Audio System: Qianyi (Rose) Sun, Scarlett Shi
Created by PhantaForma